Re: [PATCH v2 2/6] RISC-V: Add a syscall for HW probing

From: Conor Dooley
Date: Wed Feb 15 2023 - 16:11:01 EST


heh, this came in right as I went to check out by branch with this on it
and look at the rest of the series.

On Wed, Feb 15, 2023 at 12:49:35PM -0800, Evan Green wrote:
> On Tue, Feb 14, 2023 at 3:51 PM Conor Dooley <conor@xxxxxxxxxx> wrote:
> > On Mon, Feb 06, 2023 at 12:14:51PM -0800, Evan Green wrote:

> > > +On success 0 is returned, on failure a negative error code is returned.
> > > +
> > > +The following keys are defined:
> > > +
> > > +* :RISCV_HWPROBE_KEY_MVENDORID:: Contains the value of :mvendorid:, as per the
> > > + ISA specifications.
> >
> > "per the ISA specifications" sounds like dangerous wording to me! ;)
>
> I can replace "per the ISA specifications" with "as defined by the
> RISC-V privileged architecture specification" to try and make that
> more crisp.

Meh was a comment about not trusting the ISA specs, not an attempt to
be a pedant!

> > > +#define RISCV_HWPROBE_KEY_MVENDORID 0
> > > +#define RISCV_HWPROBE_KEY_MARCHID 1
> > > +#define RISCV_HWPROBE_KEY_MIMPID 2
> > > +/* Increase RISCV_HWPROBE_MAX_KEY when adding items. */
> >
> > Can't wait for that to get forgotten!
>
> I know. I could add an if (pair->key > RISCV_HWPROBE_MAX_KEY) goto
> unrecognized_key, with a label at the default switch case, which would
> effectively be a runtime guard against it. I opted not to as it's
> aesthetically harsh, but anyone can holler if they want it.

The other question to ask is, do we need RISCV_HWPROBE_MAX_KEY?
What's it for?

> > > diff --git a/arch/riscv/include/uapi/asm/unistd.h b/arch/riscv/include/uapi/asm/unistd.h
> > > index 73d7cdd2ec49..37d47302322a 100644
> > > --- a/arch/riscv/include/uapi/asm/unistd.h
> > > +++ b/arch/riscv/include/uapi/asm/unistd.h
> > > @@ -43,3 +43,11 @@
> > > #define __NR_riscv_flush_icache (__NR_arch_specific_syscall + 15)
> > > #endif
> > > __SYSCALL(__NR_riscv_flush_icache, sys_riscv_flush_icache)
> > > +
> > > +/*
> > > + * Allows userspace to probe
> >
> > That comment looks chopped off midway through.
>
> Whoops yes I

If you could flesh it

Attachment: signature.asc
Description: PGP signature